Endor was certainly not a typical battle, but neither was the Clone Wars Battle of Coruscant- and I wonder if there isn't a fleet in being thing happening here. The numbers would seem to suggest that ships are more important than planets; that it's not worth defending a planet if you lose a frigate or better doing it.
Twenty- five thousand (minimum, usual caveats apply) line destroyers- that is to say destroyers that count as a complete line in the order of battle in their own right, nominally Imperator/Imperial class, and counting the variant types I reckon you could add another fifteen thousand to that- and at least a hundred thousand, probably more, light destroyers, Victory-class and the others in the same weight class, as divided across fifty-one million inhabited worlds minimum.
Lose the naval force, and you've lost an enormous proportion of your ability to affect trade between the worlds, and to attack and defend them directly. I know, pot and kettle, analogies, but this really is sounding like the century or so on either side of the Thirty Years War- small professional forces only, much rarer and more important than any non- capital target, and the object of combat is to kill as much of the enemy's fighting strength as possible beause then the territory is ripe.
So, normally, battle is only risked when the odds are or can be manufactured to be in your favour, and potentially losing fights are things to be run away from. I think it may be that only the freak cases like Endor and Coruscant actually result in large pitched battles, when the weaker side has to stand and fight or suffer strategic disaster.
Normative combat may be a running fight, it being fairly obvious to the stronger and weaker party who is which, the weaker side attempting to disengage through hyperspace before they can be destroyed, the stronger party trying to kill the weaker before they can succeed.
At least, thus is theory before pride, stupidity, higher orders and Murphy get involved, and no doubt there are many battles fought that should not have been and more than a few that should not have been but were, but this is the model the respective abundance of ships and planets suggests to me.
In the light of that, and of the false intelligence Palpatine planted about the relatively thin defences of the construction site, the almost kamikaze aspect of Ackbar's mission makes more sense. They were expecting to have to defend against arriving Imperial reinforcements, engage and keep off the fighter- carrying ships that could have spoiled the rebel fighters' precision strike.
Ackbar, not being a fool, made the decision to go ahead with the strike after spotting the Imperial fleet and accept the losses that would be involved, heavy though they would likely be, because the strategic objective was worth the price.
Equally, withdrawing after the death star opened fire would have made hard practical sense. If it was operational to that degree, the primary objective was probably no longer feasible- the sacrifice of ships and beings would be for nothing. Why he listened to Lando is the real mystery.
"I beseech thee, In the bowels of God, think it possible that you might be wrong."
-Oliver Cromwell to Parliament, 1647
"It is good to keep an open mind; but not so open that your brains fall out." Attributed to James Oberg